Introduction to Virtual Machine Checkpoints
Backup, Upgrade, and Recover Qlik Sense
Overview
We’ll start backup process with an incredibly useful feature that is available in virtual machines - checkpoints! Although checkpoints on virtual machines are not technically backups, it can be argued that they are a form of backups.
Checkpoints serve as an incredibly convenient safety net. Checkpoints are fantastic option to have in your repertoire of tools for times when you’re upgrading Qlik Sense or installing Windows Updates or running some tests on your server. A checkpoint provides a convenient and quick way to recover from issues that may come up as you’re upgrading or experimenting with your Qlik Sense server. I LOVE checkpoints and use them often, just not for backups.
In this and in the next couple of guides, I will show you how to create a checkpoint for a virtual machine in Hyper-V Manager and how to use the checkpoints to roll back a Qlik Sense server to a stable state if/when something goes wrong.
Not Using Virtual Machines?
I do realize that you may not be running your Qlik Sense server on a virtual machine. If that’s the case, this guide is not applicable to you, and you may skip it. Although if you are considering migrating your Qlik Sense server infrastructure to a virtual environment, seeing how incredibly useful checkpoints are might give you a compelling reason to switch.
2
POINTS
1
TAKEAWAY
MED
COMPLEXITY
01. What are Checkpoints?
Checkpoints are used to create a snapshot of your server that you can use as a restore point if/when something goes wrong. They’re kind of like a sophisticated undo button. Checkpoints are not backups!
When you create a checkpoint on a virtual machine, you essentially make a copy of your entire server - all of its configurations, all software that’s installed on it, every single file, etc. If you create a checkpoint on your server when it is on and running, you will even capture the applications that are running at the time when checkpoint is created. Here’s an example of the checkpoints I’ve created on one of my Qlik Sense servers:
I just finished upgrading my Qlik Sense server to November 2024, patch 4 version and after that I’ve updated my PostgreSQL server to version 16.6. Before I started the process of upgrading PostgreSQL I took a checkpoint of my server in a healthy state with the old version of PostgreSQL. That’s the Upgraded to Nov 2024 Patch 04 checkpoint:
The idea is that if something were to go wrong during the upgrade of PostgreSQL, I would simply apply the checkpoint named Upgraded to Nov 2024 Patch 04 checkpoint to rollback my server to functioning state with older version of PostgreSQL. Checkpoints are magic.
Think of checkpoints as a magical way to freeze your server in time, a way for you to create a point in time that you can return to whenever you want. That’s the idea behind checkpoints. They allow you to capture a point in time when your server is healthy, a point that you can return to if whatever you’re about to do with your server might not go well. Like I said earlier, it’s good to think of checkpoints as a sophisticated undo button.
02. How are Checkpoints Different from Backups?
It’s tough to say and easy to argue but I will do my best to give a relevant and clear answer in the context of a Qlik Sense server.
Scenario 1: Checkpoint Only
There are two layers to virtual machines - the virtual machine itself and the host machine on which the virtual machine is running. Checkpoints provide protection against things going wrong inside of a VM. They’re no good against safeguarding issues that may happen with a host PC.
Let’s imagine a real situation. You’ve created a snapshot of the VM where your Qlik Sense server is installed and the hard drive of the host machine where the VM is running fails beyond recovery. What do you do? Nothing. The host machine where the VM was installed is gone. You’re out of luck. You start from scratch. Everything is lost. There’s no undoing that.
Scenario 2: Checkpoint & Backups
Now, same scenario but in addition to taking a snapshot of the VM, you also created backups of your Qlik Sense server and stored the backups in multiple locations. When the hard drive of the host machine that’s running your Qlik Sense server VM fails beyond recovery, what do you do? You spin up a new host machine, install and configure a virtual machine on the new host machine, and recover your Qlik Sense site from backups you’ve created. Once that’s done, you go forth like nothing happened.
Not only that, but backups can be used to restore your Qlik Sense site on both virtual and physical machines. If you don’t have a server on standby configured for virtualization, you can still use Qlik Sense backups to recover your Qlik Sense site on a dedicated, physical server.
Backups of VMs?
Now, can you take backups of virtual machines so that you would be able to restore a virtual machine on a new host PC in case something happens with the original host PC? Yes, you can, and you should. But now we’re talking about backups and not checkpoints.
The Difference
That’s the difference in a nutshell - backups are more flexible and more resilient. Fundamentally, a checkpoint is meant to be used as a quick undo mechanism, to rollback an upgrade for example. Backups, however, are meant to serve well in disaster recovery and server migration scenarios.
Guide
Summary
All Set!
It’s good to be aware of the difference between checkpoints and backups and not be confused by them. Key takeaway is to use checkpoints as a temporary safety net on top of backups to provide a quick undo and a rollback option whenever you’re testing things or running upgrades.
With that said, let’s see how to create a checkpoint, restore the server to a given checkpoint, and cleanup checkpoints when you’re done experimenting and are ready to resume regular server operations. In the next lesson we’ll go over the steps to create a checkpoint. I’ll see you there!
References
Up Next
Create a Checkpoint
Copyright © 2023 howdash LLC